Crowdsourcing Based Fuzzy Information Enrichment of Tourist Spot Recommender Systems

Abstract

Tourist Spot Recommender Systems TSRS help users to find the interesting locations/spots in vicinity based on their preferences. Enriching the list of recommended spots with contextual information such as right time to visit, weather conditions, traffic condition, right mode of transport, crowdedness, security alerts etc. may further add value to the systems. This paper proposes the concept of information enrichment for a tourist spot recommender system. Proposed system works in collaboration with a Tourist Spot Recommender System, takes the list of spots to be recommended to the current user and collects the current contextual information for those spots. A new score/rank is computed for each spot to be recommender based on the recommender's rank and current context and sent back to the user. Contextual information may be collected by several techniques such as sensors, collaborative tagging folksonomy, crowdsourcing etc. This paper proposes an approach for information enrichment using just in time location aware crowdsourcing. Location aware crowdsourcing is used to get current contextual information about a spot from the crowd currently available at that spot. Most of the contextual parameters such as traffic conditions, weather conditions, crowdedness etc. are fuzzy in nature and therefore, fuzzy inference is proposed to compute a new score/rank, with each recommended spot. The proposed system may be used with any spot recommender system, however, in this work a personalized tourist spot recommender system is considered as a case for study and evaluation. A prototype system has been implemented and is evaluated by 104 real users.